ó
õ^Rc @ s© d Z d d l m Z e d d d [ d d l Z d d l m Z d d g Z d e j f d
YZ d e j f d YZ
e d Z e
d
k r„ e n d S( sv HTML 2.0 parser.
See the HTML 2.0 specification:
http://www.w3.org/hypertext/WWW/MarkUp/html-spec/html-spec_toc.html
i’’’’( t warnpy3ks1 the htmllib module has been removed in Python 3.0t
stackleveli N( t AS_ISt
HTMLParsert HTMLParseErrorc B s e Z d Z RS( s3 Error raised when an HTML document can't be parsed.( t __name__t
__module__t __doc__( ( ( s /usr/lib/python2.7/htmllib.pyR s c B s3 e Z d Z d d l m Z d d Z d Z d Z d Z d Z d Z
d
Z d Z d Z
d
Z d Z d Z d Z d Z d Z d Z d Z d Z d Z d Z d Z d Z d Z d Z d Z d Z d Z d Z d Z! d! Z" d" Z# d# Z$ d$ Z% d% Z& d& Z' d' Z( d( Z) d) Z* d* Z+ d+ Z, d, Z- d- Z. d. Z/ d/ Z0 d0 Z1 d1 Z2 d2 Z3 d3 Z4 d4 Z5 d5 Z6 d6 Z7 d7 Z8 d8 Z9 d9 Z: d: Z; d; Z< d< Z= d= Z> d d> Z? d? Z@ d@ ZA dA ZB dB ZC dC ZD dD ZE dE ZF dF ZG dG ZH dH ZI dI ZJ dJ ZK dK ZL dL ZM dM ZN dN ZO dO ZP dP ZQ dQ ZR dR ZS dS ZT dT ZU dU ZV dV ZW dW ZX dX ZY dY ZZ dZ Z[ RS([ sĢ This is the basic HTML parser class.
It supports all entity names required by the XHTML 1.0 Recommendation.
It also defines handlers for all HTML 2.0 and many HTML 3.0 and 3.2
elements.
i’’’’( t
entitydefsi c C s t j j | | | | _ d S( s Creates an instance of the HTMLParser class.
The formatter parameter is the formatter instance associated with
the parser.
N( t sgmllibt
SGMLParsert __init__t formatter( t selfR t verbose( ( s /usr/lib/python2.7/htmllib.pyR " s c C s t | d S( N( R ( R
t message( ( s /usr/lib/python2.7/htmllib.pyt error, s c C s\ t j j | d | _ d | _ d | _ d | _ d | _ g | _ d | _
g | _ d S( Ni ( R R
t resett Nonet savedatat isindext titlet baset anchort
anchorlistt nofillt
list_stack( R
( ( s /usr/lib/python2.7/htmllib.pyR / s c C sR | j d k r" | j | | _ n, | j r> | j j | n | j j | d S( N( R R R R t add_literal_datat add_flowing_data( R
t data( ( s /usr/lib/python2.7/htmllib.pyt handle_data? s
c C s
d | _ d S( sź Begins saving character data in a buffer instead of sending it
to the formatter object.
Retrieve the stored data via the save_end() method. Use of the
save_bgn() / save_end() pair may not be nested.
t N( R ( R
( ( s /usr/lib/python2.7/htmllib.pyt save_bgnJ s c C s7 | j } d | _ | j s3 d j | j } n | S( sH Ends buffering character data and returns all data saved since
the preceding call to the save_bgn() method.
If the nofill flag is false, whitespace is collapsed to single
spaces. A call to this method without a preceding call to the
save_bgn() method will raise a TypeError exception.
t N( R R R t joint split( R
R ( ( s /usr/lib/python2.7/htmllib.pyt save_endT s
c C s) | | _ | j r% | j j | n d S( s} This method is called at the start of an anchor region.
The arguments correspond to the attributes of the tag with
the same names. The default implementation maintains a list of
hyperlinks (defined by the HREF attribute for tags) within
the document. The list of hyperlinks is available as the data
attribute anchorlist.
N( R R t append( R
t hreft namet type( ( s /usr/lib/python2.7/htmllib.pyt
anchor_bgne s
c C s3 | j r/ | j d t | j d | _ n d S( sŲ This method is called at the end of an anchor region.
The default implementation adds a textual footnote marker using an
index into the list of hyperlinks created by the anchor_bgn()method.
s [%d]N( R R t lenR R ( R
( ( s /usr/lib/python2.7/htmllib.pyt
anchor_ends s c G s | j | d S( s This method is called to handle images.
The default implementation simply passes the alt value to the
handle_data() method.
N( R ( R
t srct altt args( ( s /usr/lib/python2.7/htmllib.pyt handle_image s c C s d S( N( ( R
t attrs( ( s /usr/lib/python2.7/htmllib.pyt
start_html s c C s d S( N( ( R
( ( s /usr/lib/python2.7/htmllib.pyt end_html s c C s d S( N( ( R
R0 ( ( s /usr/lib/python2.7/htmllib.pyt
start_head s c C s d S( N( ( R
( ( s /usr/lib/python2.7/htmllib.pyt end_head s c C s d S( N( ( R
R0 ( ( s /usr/lib/python2.7/htmllib.pyt
start_body s c C s d S( N( ( R
( ( s /usr/lib/python2.7/htmllib.pyt end_body s c C s | j d S( N( R ( R
R0 ( ( s /usr/lib/python2.7/htmllib.pyt start_title s c C s | j | _ d S( N( R$ R ( R
( ( s /usr/lib/python2.7/htmllib.pyt end_title s c C s3 x, | D]$ \ } } | d k r | | _ q q Wd S( NR& ( R ( R
R0 t at v( ( s /usr/lib/python2.7/htmllib.pyt do_base s c C s
d | _ d S( Ni ( R ( R
R0 ( ( s /usr/lib/python2.7/htmllib.pyt
do_isindex” s c C s d S( N( ( R
R0 ( ( s /usr/lib/python2.7/htmllib.pyt do_link¤ s c C s d S( N( ( R
R0 ( ( s /usr/lib/python2.7/htmllib.pyt do_meta§ s c C s d S( N( ( R
R0 ( ( s /usr/lib/python2.7/htmllib.pyt do_nextidŖ s c C s$ | j j d | j j d d S( Ni t h1i ( R@ i i i ( R t
end_paragrapht push_font( R
R0 ( ( s /usr/lib/python2.7/htmllib.pyt start_h1± s c C s! | j j d | j j d S( Ni ( R RA t pop_font( R
( ( s /usr/lib/python2.7/htmllib.pyt end_h1µ s c C s$ | j j d | j j d d S( Ni t h2i ( RF i i i ( R RA RB ( R
R0 ( ( s /usr/lib/python2.7/htmllib.pyt start_h2¹ s c C s! | j j d | j j d S( Ni ( R RA RD ( R
( ( s /usr/lib/python2.7/htmllib.pyt end_h2½ s c C s$ | j j d | j j d d S( Ni t h3i ( RI i i i ( R RA RB ( R
R0 ( ( s /usr/lib/python2.7/htmllib.pyt start_h3Į s c C s! | j j d | j j d S( Ni ( R RA RD ( R
( ( s /usr/lib/python2.7/htmllib.pyt end_h3Å s c C s$ | j j d | j j d d S( Ni t h4i ( RL i i i ( R RA RB ( R
R0 ( ( s /usr/lib/python2.7/htmllib.pyt start_h4É s c C s! | j j d | j j d S( Ni ( R RA RD ( R
( ( s /usr/lib/python2.7/htmllib.pyt end_h4Ķ s c C s$ | j j d | j j d d S( Ni t h5i ( RO i i i ( R RA RB ( R
R0 ( ( s /usr/lib/python2.7/htmllib.pyt start_h5Ń s c C s! | j j d | j j d S( Ni ( R RA RD ( R
( ( s /usr/lib/python2.7/htmllib.pyt end_h5Õ s c C s$ | j j d | j j d d S( Ni t h6i ( RR i i i ( R RA RB ( R
R0 ( ( s /usr/lib/python2.7/htmllib.pyt start_h6Ł s c C s! | j j d | j j d S( Ni ( R RA RD ( R
( ( s /usr/lib/python2.7/htmllib.pyt end_h6Ż s c C s | j j d d S( Ni ( R RA ( R
R0 ( ( s /usr/lib/python2.7/htmllib.pyt do_pć s c C s@ | j j d | j j t t t d f | j d | _ d S( Ni ( R RA RB R R ( R
R0 ( ( s /usr/lib/python2.7/htmllib.pyt start_preę s c C s: | j j d | j j t d | j d | _ d S( Ni i ( R RA RD t maxR ( R
( ( s /usr/lib/python2.7/htmllib.pyt end_preė s
c C s | j | | j d d S( Nt xmp( RV t
setliteral( R
R0 ( ( s /usr/lib/python2.7/htmllib.pyt start_xmpš s
c C s | j d S( N( RX ( R
( ( s /usr/lib/python2.7/htmllib.pyt end_xmpō s c C s | j | | j d d S( Nt listing( RV RZ ( R
R0 ( ( s /usr/lib/python2.7/htmllib.pyt
start_listing÷ s
c C s | j d S( N( RX ( R
( ( s /usr/lib/python2.7/htmllib.pyt end_listingū s c C s0 | j j d | j j t d t t f d S( Ni i ( R RA RB R ( R
R0 ( ( s /usr/lib/python2.7/htmllib.pyt
start_addressž s c C s! | j j d | j j d S( Ni ( R RA RD ( R
( ( s /usr/lib/python2.7/htmllib.pyt end_address s c C s$ | j j d | j j d d S( Ni t
blockquote( R RA t push_margin( R
R0 ( ( s /usr/lib/python2.7/htmllib.pyt start_blockquote s c C s! | j j d | j j d S( Ni ( R RA t
pop_margin( R
( ( s /usr/lib/python2.7/htmllib.pyt end_blockquote
s c C sA | j j | j | j j d | j j d d d g d S( Nt ult *i ( R RA R Rc R% ( R
R0 ( ( s /usr/lib/python2.7/htmllib.pyt start_ul s c C s; | j r | j d =n | j j | j | j j d S( Ni’’’’( R R RA Re ( R
( ( s /usr/lib/python2.7/htmllib.pyt end_ul s
c C sk | j j d | j rH | j d \ } } } } | d | d <} n d \ } } | j j | | d S( Ni i’’’’i i Rh ( Rh i ( R RA R t add_label_data( R
R0 t dummyt labelt countert top( ( s /usr/lib/python2.7/htmllib.pyt do_li s c C s | j j | j | j j d d } xH | D]@ \ } } | d k r1 t | d k rh | d } n | } q1 q1 W| j j d | d g d S( Nt ols 1.R( i t .i ( R RA R Rc R* R% ( R
R0 Rm R9 R: ( ( s /usr/lib/python2.7/htmllib.pyt start_ol# s
c C s; | j r | j d =n | j j | j | j j d S( Ni’’’’( R R RA Re ( R
( ( s /usr/lib/python2.7/htmllib.pyt end_ol- s
c C s | j | d S( N( Ri ( R
R0 ( ( s /usr/lib/python2.7/htmllib.pyt
start_menu2 s c C s | j d S( N( Rj ( R
( ( s /usr/lib/python2.7/htmllib.pyt end_menu5 s c C s | j | d S( N( Ri ( R
R0 ( ( s /usr/lib/python2.7/htmllib.pyt start_dir8 s c C s | j d S( N( Rj ( R
( ( s /usr/lib/python2.7/htmllib.pyt end_dir; s c C s- | j j d | j j d d d g d S( Ni t dlR i ( R RA R R% ( R
R0 ( ( s /usr/lib/python2.7/htmllib.pyt start_dl> s c C s' | j d | j r# | j d =n d S( Ni i’’’’( t ddpopR ( R
( ( s /usr/lib/python2.7/htmllib.pyt end_dlB s
c C s | j d S( N( R{ ( R
R0 ( ( s /usr/lib/python2.7/htmllib.pyt do_dtF s c C s7 | j | j j d | j j d d d g d S( Nt ddR i ( R{ R Rc R R% ( R
R0 ( ( s /usr/lib/python2.7/htmllib.pyt do_ddI s
c C sQ | j j | | j rM | j d d d k rM | j d =| j j qM n d S( Ni’’’’i R~ ( R RA R Re ( R
t bl( ( s /usr/lib/python2.7/htmllib.pyR{ N s
c C s | j | d S( N( t start_i( R
R0 ( ( s /usr/lib/python2.7/htmllib.pyt
start_citeY s c C s | j d S( N( t end_i( R
( ( s /usr/lib/python2.7/htmllib.pyt end_citeZ s c C s | j | d S( N( t start_tt( R
R0 ( ( s /usr/lib/python2.7/htmllib.pyt
start_code\ s c C s | j d S( N( t end_tt( R
( ( s /usr/lib/python2.7/htmllib.pyt end_code] s c C s | j | d S( N( R ( R
R0 ( ( s /usr/lib/python2.7/htmllib.pyt start_em_ s c C s | j d S( N( R ( R
( ( s /usr/lib/python2.7/htmllib.pyt end_em` s c C s | j | d S( N( R
( R
R0 ( ( s /usr/lib/python2.7/htmllib.pyt start_kbdb s c C s | j d S( N( R ( R
( ( s /usr/lib/python2.7/htmllib.pyt end_kbdc s c C s | j | d S( N( R
( R
R0 ( ( s /usr/lib/python2.7/htmllib.pyt
start_sampe s c C s | j d S( N( R ( R
( ( s /usr/lib/python2.7/htmllib.pyt end_sampf s c C s | j | d S( N( t start_b( R
R0 ( ( s /usr/lib/python2.7/htmllib.pyt start_strongh s c C s | j d S( N( t end_b( R
( ( s /usr/lib/python2.7/htmllib.pyt
end_strongi s c C s | j | d S( N( R ( R
R0 ( ( s /usr/lib/python2.7/htmllib.pyt start_vark s c C s | j d S( N( R ( R
( ( s /usr/lib/python2.7/htmllib.pyt end_varl s c C s | j j t d t t f d S( Ni ( R RB R ( R
R0 ( ( s /usr/lib/python2.7/htmllib.pyR p s c C s | j j d S( N( R RD ( R
( ( s /usr/lib/python2.7/htmllib.pyR r s c C s | j j t t d t f d S( Ni ( R RB R ( R
R0 ( ( s /usr/lib/python2.7/htmllib.pyR u s c C s | j j d S( N( R RD ( R
( ( s /usr/lib/python2.7/htmllib.pyR w s c C s | j j t t t d f d S( Ni ( R RB R ( R
R0 ( ( s /usr/lib/python2.7/htmllib.pyR
z s c C s | j j d S( N( R RD ( R
( ( s /usr/lib/python2.7/htmllib.pyR | s c C s d } d } d } xe | D]] \ } } | j } | d k rF | } n | d k r[ | } n | d k r | j } q q W| j | | | d S( NR R& R' R( ( t stript lowerR) ( R
R0 R& R' R( t attrnamet value( ( s /usr/lib/python2.7/htmllib.pyt start_a s c C s | j d S( N( R+ ( R
( ( s /usr/lib/python2.7/htmllib.pyt end_a s c C s | j j d S( N( R t add_line_break( R
R0 ( ( s /usr/lib/python2.7/htmllib.pyt do_br s c C s | j j d S( N( R t add_hor_rule( R
R0 ( ( s /usr/lib/python2.7/htmllib.pyt do_hr s c
C s d } d } d } d } d } d } xĪ | D]Ę \ } } | d k rL | } n | d k ra | } n | d k rv | } n | d k r | } n | d k r¾ y t | } Wq¾ t k
rŗ q¾ Xn | d k r+ y t | } Wqń t k
rķ qń Xq+ q+ W| j | | | | | | d S(
NR s (image)i t alignR- t ismapR, t widtht height( t intt
ValueErrorR/ (
R
R0 R R- R R, R” R¢ R R ( ( s /usr/lib/python2.7/htmllib.pyt do_img s4
c C s | j | | j d S( N( RV t
setnomoretags( R
R0 ( ( s /usr/lib/python2.7/htmllib.pyt do_plaintext¶ s
c C s d S( N( ( R
t tagR0 ( ( s /usr/lib/python2.7/htmllib.pyt unknown_starttag¼ s c C s d S( N( ( R
RØ ( ( s /usr/lib/python2.7/htmllib.pyt unknown_endtagæ s (\ R R R t htmlentitydefsR R R R R R R$ R) R+ R/ R1 R2 R3 R4 R5 R6 R7 R8 R; R< R= R> R? RC RE RG RH RJ RK RM RN RP RQ RS RT RU RV RX R[ R\ R^ R_ R` Ra Rd Rf Ri Rj Rp Rs Rt Ru Rv Rw Rx Rz R| R} R R{ R R R R R R R R R R R R R R R R R R R
R R R R R R„ R§ R© RŖ ( ( ( s /usr/lib/python2.7/htmllib.pyR s²
c C sA d d l } d d l } | s. | j d } n | oA | d d k } | rT | d =n | rg | d } n d } | d k r
| j } nC y t | d } Wn- t k
rĒ } | Gd G| GH| j d n X| j } | | j k rš | j n | r| j } n | j
| j } t | } | j
| | j d S( Ni’’’’i i s -ss test.htmlt -t rt :( t sysR t argvt stdint opent IOErrort exitt readt closet
NullFormattert AbstractFormattert
DumbWriterR t feed( R. RÆ R t silentt filet ft msgR t p( ( s /usr/lib/python2.7/htmllib.pyt testĆ s2
t __main__( R t warningsR R R R t __all__t SGMLParseErrorR R
R R RĄ R ( ( ( s /usr/lib/python2.7/htmllib.pyt s ’ '